Factors associated with incongruence between patient and carer reports in advanced COPD

Background: Carer assessment of patient symptoms and care is critical for decision-making by themselves, clinicians and researchers. However few studies have examined proxy reporting validity in COPD or the primary care setting. Aims: To determine agreement between patient and carer report on symptoms and support for activities of daily living (ADL), and factors associated with incongruence. Methods: A prospective population-based study of 119 patients with advanced COPD and their carers. Wilcoxon signed rank tests and ICC determined group level bias and individual level agreement respectively. Chi-square, one-way ANOVA and Kruskal-Wallis tests identified factors associated with incongruence. Results: There was no group-level bias for any symptoms, but carers estimated greater support for cooking (p=0.012), housework (p=0.021) and paperwork (p=0.020). Individual-level agreement was moderate for constipation (ICC=0.490), diarrhoea (ICC=0.461) and depression (ICC=0.445), fair for anxiety (ICC=0.372), fatigue (ICC=0.326) and dyspnoea (ICC=0.231), moderate for personal ADLs and fair for practical ADLs. Greater estimation by carers was associated with non-spousal relationship, non-cohabitation, more hours caring/week, impact on social life and finances and unmet carer support needs, and greater estimation by patients with longer duration of COPD. Bidirectional incongruence was associated with more exacerbations and patient and carer physical and psychological co-morbidities. Conclusion: Patient-carer agreement was worse for less observable symptoms and ADLs, less patient-carer interaction, greater disease and caring burden, and poorer dyad health. Greater communication within dyads should be promoted.